Roof 3D modeling

From OpenStreetMap Wiki
Jump to: navigation, search

The 3D modeling techniques described in this page are based on S3DB definition and Kendzi 3D PlugIn for JOSM.

1. Roof with one slope this very common but for the most users complicated roof needs in addition to roof:shape=skillion tag still the rotation definition- means the ftom top to down direction of fall of the water on the roof. This angle is described as roof:slope:direction=<value in degree>

An example shows how to easily determine this angle and what could be done wrong:

1. An existing building=yes with roof:shape=hipped shape: UseofRoof SlopeDirection1.JPG


The so called skeleton algorithm generates following 2d result: UseofRoof SlopeDirection2.JPG

And as 3D model we would get this:

UseofRoof SlopeDirection3.JPG

Our wishful result should look different:

UseofRoof SlopeDirectio4.JPG


This can be realized only by use of tag roof:slope:direction for each surface:

UseofRoof SlopeDirection4.JPG

Measurement of roof:slope:direction

This method need in JOSM PlugIn: BuildingTools

Activate the wishful roof polygon and add the attributtes: roof:shape=skillion + roof:slope:direction=?

1. Klick on the BuildingTools symbol 2. start to draw the building (from the lowest roof polygon corner) using building tools. 3. Note the value of angle shown bottom left in JOSM 4. Put this angle als value of roof:slope:direction tag.

MarekRoofSlopeDirectionMeasurement.jpg

enjoy the modeling!

Sphere modeling

I know from Wikipedia, the height of the tower in "my" building is 80m I assume, the middle of the tower is 79.30m.

Because there is no sphere in the specification, I construct a sheet model:

400px this sketch help me to avoid wrong height values:

MarekModelingAsphere1.jpg

Sheet nr. 1.:

MarekModelingAsphere2.jpg

Sheet nr. 2.:

MarekModelingAsphere3.jpg

Sheet nr. 3.:

MarekModelingAsphere4.jpg

I decide, this looks not good enough and add one sheet more:

MarekModelingAsphere5.jpg

The result now:

MarekModelingAsphereResult.jpg

Complex shapes

The approach is easy to understand: Try to divide complicated geometries in simple elements. Similar to cone modeling. The difference to, let´s say Lego elements, is the possibility of spatial overlapping,fusion of elements.

Complex3dShape.png

results in:

Complex3dShape3dResult.JPG